In February 2026, OpenAI launched ads in ChatGPT while Perplexity walked away from advertising entirely. The honest reason behind that split, not just principle, tells you exactly which AI engines you can actually buy your way into, and which ones you have to earn.
Two AI companies made opposite bets on advertising in the same month, and the difference between them tells you something concrete about where to put your effort. In February 2026, OpenAI began testing ads inside ChatGPT, for users on the Free and Go tiers only, clearly labeled, kept separate from the answer itself, with Plus, Pro, Business, and Enterprise users staying ad-free. That test expanded fast, a self-serve Ads Manager followed by May, no minimum spend required, live in the US, UK, Japan, South Korea, Brazil, and Mexico within months.
Perplexity went the other way, and the timing wasn't a coincidence. The company had been testing sponsored placements since late 2024, quietly stopped signing new advertisers in October 2025, and by February 2026 confirmed the retreat was permanent. Executives told the Financial Times the concern was user trust, an answer engine doesn't have a separate organic zone for a skeptical user to retreat to the way a search results page does. If the one answer you're given might be paid for, the whole product's credibility is what's on the line, not just that one result.
The trust explanation is real, but it isn't the whole story, and the honest version includes the part that's less flattering. Perplexity's ad product reportedly struggled commercially before the principled exit, advertisers found the targeting and measurement tools thin, and the company was working with a user base of roughly 60 million monthly users against ChatGPT's several hundred million weekly actives. At that gap in scale, ad economics are genuinely difficult to make work regardless of how the product is positioned. Both things can be true at once, a real trust concern, and a product that wasn't going to win the ad business anyway.
This matters less as a story about two companies and more as a map of where your money and effort actually work. Right now, there are exactly two AI surfaces where you can buy placement directly, Google AI Mode and ChatGPT. Everywhere else, Perplexity, Claude, and by extension any engine that follows Perplexity's lead, there's nothing to purchase. The only way in is being cited, which is earned through content structure and authority, not spend.
That split should change how a founder budgets attention, not just ad dollars. Trust data backs up why some AI platforms are choosing to stay ad-free as a selling point, one widely cited industry survey found a majority of users say ads inside an AI answer reduce their trust in that answer. Anthropic has made the same bet as Perplexity, keeping Claude ad-free deliberately. That's not a neutral design choice, it's a wager that some meaningful share of users will actively choose the platform that doesn't blend commerce into its answers.
The practical takeaway isn't picking a side in that argument. It's building two different muscles at once, because the AI landscape now genuinely requires both. On ChatGPT and Google AI Mode, there's a real, buyable channel worth testing if it fits your funnel, alongside the separate work of earning organic citation there too, since as covered in an earlier piece here, buying the ad slot doesn't buy the citation, they're tracked and won separately even on the platforms that sell both. On Perplexity and Claude, there's only one lever, being the source that's actually cited, which means the structural work, direct answers, clear sourcing, current information, isn't optional there. It's the entire strategy.
Worth checking directly rather than assuming: open Perplexity and Claude and search your own top questions, the ones your actual customers would ask before buying. See whether you show up as a cited source at all right now. If you don't, that's not a budget problem you can fix with an ad campaign, on those platforms there's no ad campaign to run. It's a content and authority problem, and it's worth knowing which kind of gap you're actually looking at before deciding where the next quarter's effort goes.
The AI visibility landscape is genuinely splitting in two right now, worth having someone tracking which half your business actually needs.
Talk to us →© 2026 The Chill Pill